Radial basis function interpolation of fields resulting from nonlinear simulations

Abstract Three approaches for construction of a surrogate model result field consisting multiple physical quantities are presented. The first approach uses direct interpolation the space on input space. In second and third Singular Value Decomposition is used to reduce size. reduced order models, amplitudes corresponding different basis vectors interpolated. A quality measure that takes into account parts defined. As very cheap evaluate, it can be efficiently optimize hyperparameters all models. Based measure, criterion proposed choose number performance models resulting from three compared using based validation set. It found novel effectively select vectors. choice method significantly influences model.